Redundant ProviewR project

Hello,I have a question regarding redundancy in ProviewR. I’m wondering whether it is possible to run two machines independently using the same ProviewR project, so that if system A experiences a failure, the project continues running on system B.If this is possible, how would one set it up?Thanks in advance,
/User

Hi User,

There is a description here

www.proviewr.se/doc/en_us/man_redu.pdf

/Claes

Thanks for the response.I realized that I had something misconfigured at the beginning. I do have a follow-up question, though.

I was able to make my system appear redundant, but I’m running into an issue when redundancy is combined with Modbus TCP.
I have two systems (System A and System connected to a Modbus server. Whichever system starts first is the one that successfully sends read and write requests to the Modbus server, and everything works as expected. The second system is able to connect to the server, but it does not send read or write requests while the first system is active.
When System A loses its connection due to a shutdown or failure, System B does not take over the Modbus communication. Even though System B is then considered the active system, read and write requests still do not work. If System B is started first, the opposite behavior occurs.

I was wondering what could be causing this issue or what I should check to enable proper failover of the Modbus TCP connection.Thanks in advance
/User

hello user,

I ran into that same problem, or more sort of.
At first I used IO from Beckhoff. The nodes made both a connection since the interface allow this. Only the first connected interface was allowed to read/write the date, so when the system switched over, the interface resetted it connection and the outputs went low (the inputs could be covered leaving them in the last state)..
This was the reason I switched over to Siemens ET200SP (IM155MF interface). This interface allowed me to really connect 2 modbus slaves and my problems were gone, until...
The last added hardware where modbus RTU devices. Here is where a whole mess started. Since the standby node is questioning the devices, the RTU data gets mixed, or better, the RTU devices are getting confused by the load demand of polling registers double in random order. I've got for ex. pumps with Hydrovars from Xylem and they definetly don't like it and gave me a lot of timeouts (tip: when the communication goes into timeout, the plc process will slow down to due the wait for timeouts, so split the communication into seperate PLC's).
The solution for this was disabling the communication to those devices from the standby node.
I than asked Claes to create a special feature so that the slave node is modbus polling is inhibited. Unfortunatly I haven't found the time to test this out, so the config of the slave node is still with the polling disabled. This means when I switch to my backup node there will be a problem. Fortunatly, this happened to the system only once for nog (also due communication problems).